sweet-search 2.6.0 → 2.6.1

package/core/indexing/ast-chunker.js

@@ -349,7 +349,11 @@ export class ASTChunker {
     // chunker uses tree-sitter-cpp instead of tree-sitter-c.
     const langInfo = resolveLanguage(filePath, content);
     if (!langInfo || !langInfo.chunker) {
-      return this.parseGenericFile(filePath, content);
+      // Mapped-but-chunkerless languages (e.g. Clojure: no tree-sitter grammar,
+      // and paren-delimited forms don't fit the brace/indent/endKeyword parsers)
+      // fall back to lossless generic windowing but keep their resolved language
+      // id so chunks are tagged 'clojure' rather than generic 'text'.
+      return this.parseGenericFile(filePath, content, langInfo?.id || 'text');
     }
 
     // Try tree-sitter WASM first for supported languages
@@ -909,7 +913,7 @@ export class ASTChunker {
     return subChunks;
   }
 
-  parseGenericFile(filePath, content) {
+  parseGenericFile(filePath, content, language = 'text') {
     const lines = content.split('\n');
     const chunks = [];
     const CHUNK_SIZE = 50;
@@ -921,7 +925,7 @@ export class ASTChunker {
       const chunkContent = lines.slice(start, end).join('\n');
 
       if (chunkContent.trim().length > 20) {
-        chunks.push(this.buildChunk(chunkContent, filePath, 'text', 'code', 'unknown', start, end - 1));
+        chunks.push(this.buildChunk(chunkContent, filePath, language, 'code', 'unknown', start, end - 1));
       }
 
       start = end - OVERLAP;

package/core/infrastructure/config/search.js

@@ -53,6 +53,7 @@ export const FILE_PATTERNS = {
     // Source code (all major languages)
     '**/*.{js,jsx,ts,tsx,mjs,cjs}',      // JavaScript/TypeScript
     '**/*.{java,kt,kts,scala,groovy}',    // JVM
+    '**/*.{clj,cljc,cljs,edn}',           // Clojure / ClojureScript / EDN
     '**/*.{py,pyi}',                       // Python
     '**/*.go',                              // Go
     '**/*.rs',                              // Rust

package/core/infrastructure/language-patterns/maps.js

@@ -42,6 +42,15 @@ export const EXTENSION_MAP = {
   // Scala
   '.scala': 'scala',
 
+  // Clojure / ClojureScript / EDN
+  // No tree-sitter-clojure grammar ships in the bundle and the brace-based
+  // regex chunker would fragment paren-delimited Lisp forms (and drop sub-30-char
+  // defns via MIN_CONTENT_LENGTH), so there is intentionally no LANGUAGES.clojure
+  // entry — getLanguageByExtension returns the chunker-less fallback and
+  // ast-chunker routes these through parseGenericFile (lossless 50-line windows),
+  // tagged with this 'clojure' language id rather than generic 'text'.
+  '.clj': 'clojure', '.cljc': 'clojure', '.cljs': 'clojure', '.edn': 'clojure',
+
   // Dart
   '.dart': 'dart',
 

package/package.json

@@ -1,6 +1,6 @@
 {
   "name": "sweet-search",
-  "version": "2.6.0",
+  "version": "2.6.1",
   "description": "Sweet Search - SOTA Hybrid Code Search Engine with WASM CatBoost Query Router, Semantic/Lexical/Structural Search, and Multilingual Support",
   "type": "module",
   "main": "core/search/sweet-search.js",
@@ -167,13 +167,13 @@
     "vitest": "^4.0.16"
   },
   "optionalDependencies": {
-    "@sweet-search/native-darwin-arm64": "2.6.0",
-    "@sweet-search/native-darwin-x64": "2.6.0",
-    "@sweet-search/native-linux-arm64-gnu": "2.6.0",
-    "@sweet-search/native-linux-arm64-gnu-cuda": "2.6.0",
-    "@sweet-search/native-linux-x64-gnu": "2.6.0",
-    "@sweet-search/native-linux-x64-gnu-cuda": "2.6.0",
-    "@sweet-search/bg-priority": "2.6.0"
+    "@sweet-search/native-darwin-arm64": "2.6.1",
+    "@sweet-search/native-darwin-x64": "2.6.1",
+    "@sweet-search/native-linux-arm64-gnu": "2.6.1",
+    "@sweet-search/native-linux-arm64-gnu-cuda": "2.6.1",
+    "@sweet-search/native-linux-x64-gnu": "2.6.1",
+    "@sweet-search/native-linux-x64-gnu-cuda": "2.6.1",
+    "@sweet-search/bg-priority": "2.6.1"
   },
   "engines": {
     "node": ">=18.0.0"